2 Cost function: Consider a Linear regression model in 2D space with 2 weights and no bias term: With MSE (mean-squared error), the total cost for the dataset is eo)2 Given a training set consisting of N 3 examples: o (2,0),t) 1 Write the cost in the form Q: What are the values of ci , di , C2, d2 and c3 ? [2 pts] Q: Plot the above function for e = 1; you may hand-draw it or use desmos graphing calculator, what shape does the cost function represent? [5 pts] Q: Repeat by setting e 0. What do you observe? Why? [2 pts] Q: Gradient Descent usually converges to the local minima; however, Linear Regression converges to the global minimum as there is only one minimum. What is your estimate of the number of different values of w1 and w2 at convergence? 13 pts] 1 value 1 possible minmiu value 1 local minium 2 Cost function: Consider a Linear regression model in 2D space with 2 weights and no bias term: With MSE (mean-squared error), the total cost for the dataset is eo)2 Given a training set consisting of N 3 examples: o (2,0),t) 1 Write the cost in the form Q: What are the values of ci , di , C2, d2 and c3 ? [2 pts] Q: Plot the above function for e = 1; you may hand-draw it or use desmos graphing calculator, what shape does the cost function represent? [5 pts] Q: Repeat by setting e 0. What do you observe? Why? [2 pts] Q: Gradient Descent usually converges to the local minima; however, Linear Regression converges to the global minimum as there is only one minimum. What is your estimate of the number of different values of w1 and w2 at convergence? 13 pts] 1 value 1 possible minmiu value 1 local minium